Share via


Installing Compute Cluster Pack on a Compute Node

Applies To: Windows Compute Cluster Server 2003

The Compute Cluster Pack (CCP) provides the components that configure a server as a head node or a compute node in Windows Compute Cluster Server 2003. The Compute Cluster Pack is a combination of interfaces, utilities, and management infrastructure which enable high performance computing on Windows Server operating systems.

Installing Compute Cluster Pack

Before you install the Compute Cluster Pack and create a compute node, be sure that the following prerequisites have been met:

  • The prospective node must already have Windows Server 2003, Compute Cluster Edition or a supported operating system installed.

  • The computer must be connected to at least one network, depending on your choice of cluster networking topology, and you must be a local administrator on that computer.

  • The prospective compute node must be a member of the same domain as the head node.

  • You must be a local administrator on the node.

To create a compute node

  1. Run Setup.exe from the Compute Cluster Pack CD-ROM or from the source location of this media. The Compute Cluster Pack Installation Wizard will be displayed.

  2. On the Getting Started page, verify that you want to create a compute node by clicking Next.

  3. On the Microsoft Software License Agreement page, if you accept the terms of the license agreement, select the appropriate radio button, and then click Next.

  4. On the Select Installation Type page, select Join this server to an existing compute cluster as a compute node.

  5. Enter the name of the head node, or click Browse and locate the head node. If the specified head node is not found, an error message is displayed. Click OK to dismiss the error message and enter another head node or browse to locate it. If the specified head node cannot be accessed or you do not specify a head node, the compute node will not be joined to a cluster. Optionally, select Install Microsoft Compute Cluster Pack Client Utilities on compute node also if you want to install the client utilities on a compute node. Then, click Next.

    Note

    You can add the node to a cluster at a later time with the Add Node Wizard from the Compute Cluster Administrator. Using this wizard, select the Manual Addition method to add the node to an existing cluster.

  6. Compute Cluster Pack Setup examines the compute node for Microsoft .NET Framework 2.0 and highlights it on the Install Required Components page if it is not present. Click Install to install this software.

  7. When the required software is installed on the compute node, Compute Cluster Pack is highlighted. Click Install.

  8. On the Select Installation Location page, click Next to accept the default location where the Compute Cluster Pack will be installed, or click Browse to change the folder.

  9. On the View Installation Summary page, review the installation options you have selected. Note the components that will be installed. If you are satisfied with the installation options, click Install to begin the installation of the Compute Cluster Pack. If you want to change the installation, click Previous to change it. You can cancel installation of the Compute Cluster Pack at any time by clicking Cancel.

  10. The Installation Progress page is displayed. You will be able to see the progress of the installation as each of the Compute Cluster Pack components are installed.

    1. If the installation is successful, the Completing the Wizard page is displayed. Click Finish to close the wizard.

    2. If the installation is not successful, the Wizard Not Completed page is displayed. Click Finish to close the wizard and investigate the cause of failure.

  11. When the installation is complete, the To Do List page is displayed.

Next steps

If you have both created the compute node and added it to a cluster, the new node is in a Pending Approval state. See Managing Compute Nodes for more information about approving nodes in order to begin receiving and processing jobs from users.